The ethical judiciousness of the major decisions made by the principal investigator (Dr. Togias) and the Johns Hopkins University (JHU) Institutional Review Board (IRB) with respect to the case of Ellen Roche was highly problematic.  I will focus my reflection on three issues of more general import that the case brings to mind which I believe are worthy of the attention of clinical researchers and the bioethics community:  acceptable levels of risk in research,  standards for literature searches, and the nature of collective responsibility.   

1.  Assessing Risk and Protecting Human Subjects in Therapy-Oriented Versus Investigational Science-Oriented Research:  The study under review involved an attempt to induce the equivalent of an asthma attack in a healthy, non-asthmatic research subject in order to learn about the nervous system and asthma.  The description of the experiment and its consent form stated that the experiment was not intended personally to benefit the research subject.  This is true.  But it is also the case that the drugs administered in the experiment were, themselves, under therapeutic consideration.  While the information potentially to be obtained might assist in developing future therapeutic interventions, the experiment, itself, was not directly linked to a proposed therapy or diagnosis for asthma.  It was a purely physiologic study.

For that reason, even if the hexamethonium had been FDA-approved, its experimental use and its administration via inhalation would not have been addressed.  It had been used only experimentally, not clinically, in only twenty research subjects, and those experiments were not designed to evaluate the drugs safety.  As the internal investigation by Hopkins after the study determined:

One way to obtain evidence of safety [of inhaled hexamethonium] would be to submit an IND application to the FDA and secure approval of it. It is understandable why the IRB did not ask Dr. Togias to submit an IND application for inhaled hexamethonium, given the repeated inability to obtain a written response from the FDA on this matter and the uncertainty of interpretation of the criteria for requiring submission of an IND application. Because of the relative paucity of published safety data of inhaled hexamethonium (only a total of 20 subjects studied), the committee believes that the IRB should have asked Dr. Togias to obtain an opinion from the FDA about whether the proposed use of inhaled hexamethonium required an IND number or was exempt. The committee also believes that the IRB should have requested more details from Dr. Togias about exactly how the hexamethonium was going to be prepared, stored, and tested for purity prior to administration to research subjects, although it is doubtful that more specification of these procedural details would have prevented the fatal outcome.[1]

Even if obtaining information on the drug from the FDA would have been time-consuming, it should have been required by the JHU-IRB. 

The Alliance for Human Research Protection (www.ahrp.org) reports that the Food and Drug Administration (FDA) has exempted academic research centers from obtaining FDA permission to conduct human experiments with unapproved drugs (as is required of pharmaceutical companies) because the process necessary to obtain permission is expensive.[2]  Safety should take precedence in research involving human subjects – particularly in physiologic studies.

Whereas investigations of drugs with potential therapeutic applications go through a long series of analysis and animal testing before being tested on human beings, the experiment involving Ms. Roche made use of a drug (administered in a somewhat novel way and in conjunction with another drug, methacholine) for purely experimental purposes.  Protocols that involve intentionally exposing healthy research subjects to potential harm without any deserve at least as much protection as subjects in protocols involving potentially therapeutic drugs.  Such protection should include animal testing before administration of a drug or combination of drugs to a human being.

Additionally, because there is a potential conflict of interest, investigators should not evaluate the potential significance of adverse events on their own.  The adverse reaction of the first participant (prolonged cough and shortness of breath which lasted for eight days) should have been reported immediately.  Failure to do so was irresponsible.[3] 

In light of the fact that drugs chemically related to hexamethonium had been shown to cause lung damage, that there were very limited data regarding the potential effects of inhaled hexamethonium, and that the combination of hexamethonium and methadholine had not been studied in animals or humans, approval of the protocol as presented to the JHU-IRB was ill-advised.  More generally, participants in experimental protocols deserve at least as much protection as participants in protocols that are testing for therapeutic or diagnostic potency.

2.  Literature Review Standards and Methods:  Eva Perkins, a medical librarian,[4] makes a strong case for medical centers’ being required to maintain a staff of specialized medical librarians, particularly librarians who are familiar with research involving internet searches for biomedical and pharmacological information.  A qualified medical librarian would have known how to conduct a more effective and extensive search on hexamethonium and its toxicity; moreover, he or she would have known the date limitations (most PubMed material goes back only to 1966) and the search-term conventions for various databases and periods that researchers might use.  Such conventions change over time, and a thorough familiarity with them is essential to research and safety.  Without such knowledge, researchers are likely to miss valuable, life-saving or life–threatening information.   Institutions should develop protocols for experimental design that require the consultation of individuals, including specialized research assistants and librarians, who have specialized training. 

The interesting philosophical question here is one of epistemology.  It calls to mind Donald Rumsfeld’s[5] famous quotation, “There are known knowns. There are things we know that we know. There are known unknowns. That is to say, there are things that we now know we don’t know. But there are also unknown unknowns. There are things we do not know we don’t know.”  The determination of whether one has done sufficient preliminary research is a matter of practical reasoning.  The inclusion of medical librarian/research specialists on a research team would seem to be a necessity in today’s complex world of information storage and retrieval.

3.  Institutional and Individual Moral Responsibility:  Dr. Togias, the JHU-IRB, and the Board of Trustees of Johns Hopkins University all bear moral responsibility for Ellen Roche’s death.  All violated the public trust by failing to adequately protect human research subjects.  Although, in general, the degree of moral agency exercised by each individual decreases as distance from the act increases, no one in the institutional chain of command is absolved from moral responsibility when the potential for harm could or should have been avoided but was not. 

A broader philosophical question that arises here as well:  what is the nature of corporate or institutional responsibility? 

The findings of prisoner abuse by U.S. soldiers in Abu Ghraib should have provoked reflection by our nation’s leaders on whether those involved were “a few bad apples” or whether their behavior was a manifestation of elements of a more pervasive ethos.  Similarly, the judgments of the principal investigator and of the IRB (with respect to the scientific validity and importance of the study, the adequacy of care for participant safety, the reporting of adverse events, the novel use of drugs not approved by the FDA, etc.) should have compelled all JHU institutional leaders to reflect upon whether their institution’s ethos contributed to the misjudgments that led to Ms. Roche’s death.  JHU did engage in that sort of reflection, and the institution is to be commended.  JHU has admitted responsibility and has taken steps to reduce the likelihood of similar errors in the future. The temporary suspension of research at JHU by order of the Office of Human Research Protections (OHRP) was appropriate in light of not only Ms. Roche’s death but also a number of other problematic conditions that OHRP discovered at the institution.[6]

On a more general note, events such as this should encourage all of us who work in research settings to evaluate both our own and our colleague’s attitudes toward ethics, safety, and “compliance” – which should not be conflated.  All members of an institution should be encouraged to frequently assess their work environment and with respect to how clearly focused members are not only on obtaining grants, conducting experiments, publishing, etc. but also on the larger social goals that motivate and justify research in the first place. 

The fact that the determination of what constitutes acceptable risk is a matter of practical judgment that cannot be reduced to or captured in a formula does not imply that such a judgment should be taken lightly.  Cases like this ought to encourage everyone in the research community to redouble his or her efforts to question the significance of all experiments involving human subjects and whether the risks of an experiment are as minimal as possible and are justified by the potential therapeutic, diagnostic, or physiologic information to be attained.  Institutions should attempt to establish clear standards of protection and preliminary research needed for each type of experimentation